(A) Schematic illustration of the imaging setup (top) and the fruit fly visual system around L2 cells (bottom). Inset, example of the region imaged, with six L2 terminals expressing ASAP2s. Expression of the other sensors was comparable. Scale bar, 5 μm. (B) L2 responses to alternating 300 ms dark and light flashes, as measured with different indicators. The black bar indicates the dark period and the white bar indicates the light period. Top, mean response across all cells (n = 44 cells from 3 flies for ASAP1, 111 cells from 5 flies for ASAP2s, 65 cells from 5 flies for ArcLight, 64 cells from 3 flies for MacQ-mCitrine, 23 cells from 4 flies for Ace2N-2AA-mNeon, and 232 cells from 10 flies for GCaMP6f). Each cell contributes its average response across 100 trials (one trial = 1 dark flash and one light flash). Horizontal black lines are the mean fluorescence per trial; indicators with slow or asymmetric responses to dark and light flashes can produce traces that do not begin or end at the mean fluorescence. Bottom, for ASAP1, ASAP2s, ArcLight, Ace2N-2AA-mNeon, and GCaMP6f, five exemplar single-trial responses from a single representative L2 cell (gray) and the same cell’s mean response over all trials (colored, n = 100 trials). Solid line is mean; lighter shading is ±1 SEM. Because MacQ-mCitrine didn’t produce an apparent stimulus-evoked response when averaging across all cells (top trace), we plotted the mean optical traces of individual cells (bottom traces) to illustrate that stimulus-evoked responses were also not apparent in any cells. (C) Schematic illustrating the response parameters quantified in panel D. Amax, the maximal amplitude of the fractional fluorescence change (|ΔF/F|); tpeak, the time at which Amax occurs, relative to the start of the flash; τdecay, the time constant of the decay from Amax. (D) For each voltage sensor, Amax, tpeak, and τdecay are plotted for depolarizations (left) and hyperpolarizations (right). Sample sizes are the same as in panel B, except for some measurements of τdecay that did not meet our inclusion criterion (see Materials and methods). Amax and τdecay were analyzed with the t-test, while tpeak was analyzed with the Mann-Whitney U-test. We used the Bonferroni method to correct for multiple pairwise comparisons between ASAP1, ASAP2s, and ArcLight values for a given response parameter (Amax, tpeak, and τdecay) and sign of voltage change (depolarization or hyperpolarization). *p<0.05, **p<0.01, ***p<0.001.